audit-paper

audit-paper is a command for Claude Code from TobiasBlask/open-paper-machine. It costs 26 tokens per session (300 once invoked), scanned A, original, MIT.

A command that checks whether a research paper’s claims match the code repository used to support them. TDD here means testable claims and evidence, not a programming method.

In plain words
What is it for?
It extracts claims, maps the repository, searches for supporting evidence, classifies each claim, and writes an audit report.
Why use it?
It catches claims that are confirmed only partly, missing from the code, or inconsistent with the implementation before submission.

Command for Claude Code

Written for Claude Code: $ARGUMENTS substitution.

Part of the open-academic-paper-machine plugin — 33 skills, 21 commands, 4 agents shipped together

Good fit It extracts claims, maps the repository, searches for supporting evidence, classifies each claim, and writes an audit report.

Compare 6 commands from other repositories ↓
Install with agentmods
npx agentmods add commands/tobiasblask/open-paper-machine/audit-paper
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/TobiasBlask/open-paper-machine

Made for: Claude Code.

Or install open-academic-paper-machine, the plugin that ships this one along with the rest of its 33 skills, 21 commands, 4 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for audit-paper

README.md
[![agentmods](https://agentmods.dev/badge/commands/tobiasblask/open-paper-machine/audit-paper.svg)](https://agentmods.dev/commands/tobiasblask/open-paper-machine/audit-paper)
Your own site
<a href="https://agentmods.dev/commands/tobiasblask/open-paper-machine/audit-paper"><img src="https://agentmods.dev/badge/commands/tobiasblask/open-paper-machine/audit-paper.svg" alt="Measured on agentmods" height="20"></a>
Per session 26 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 300 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00026 $0.00300
Opus 5 $0.00013 $0.00150
Sonnet 5 $0.00005 $0.00060
Haiku 4.5 $0.00003 $0.00030

Measured 8d ago against content hash e9f48540dce0, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-08, from the pricing page.

Security

Grade A, and why

audit-paper scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

commands/audit-paper.md · 25 lines

What it actually says

Audit Paper vs. Code: $ARGUMENTS

Activate the audit-engine skill. Read skills/audit-engine/SKILL.md in full.

Execute the complete 5-step audit workflow:

  1. Extract testable claims from paper.tex (or draft.md)
  2. Locate the code repository (from $ARGUMENTS, or scan paper for a code URL, or ask once)
  3. Map the repo (entry points, configs, data loaders, results)
  4. Search for evidence per claim using Grep + Read with file:line citations
  5. Classify each claim as CONFIRMED / PARTIAL / MISSING / MISMATCH / NOT_AUDITABLE
  6. Generate outputs/audit_report.md

$ARGUMENTS may contain:

  • A repo path or GitHub URL to audit against
  • A scope hint (e.g., "only Section 4", "only the training claims", "tier 1 only")
  • Both

If no code repository can be located, stop and ask the user for a path or URL before proceeding.

This is a static audit: no code is executed. For full reproducibility, re-run training and evaluation manually after the audit surfaces any MISMATCH or MISSING claims.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 8d ago First seen · 25 lines · 26 tokens per session scan A e9f48540dce0

Subscribe to this mod's changes

audit-paper is a command published in the GitHub repository TobiasBlask/open-paper-machine (18 stars, last pushed 5mo ago), licensed MIT. It adds 26 tokens to every session and 300 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.